browser-verification
Skill9 starsChanged 2 months agoMITView on GitHub
Confirms a user-interface change actually works by loading it in a real browser and checking rendering, interaction, keyboard access, console output, network calls, and responsive behavior. Use after any visual or interactive change, when a test passes but the page might not, or before claiming a UI change is done.
5 other projects use a version of this.
Adopted by
Most recently taken up by
- SWEStash/swe-workflow-skillstheir own version, first seen Nov 8, 2026
- arslanvuzmal/VoxCircuittheir own version, first seen Oct 13, 2026
- nahid-sparktales/agent-dispatchertheir own version, first seen Sep 21, 2026
- markoblogo/abvx-agent-skillstheir own version, first seen Jun 14, 2026
- motao123/dev-workflow-kittheir own version, first seen Jun 13, 2026
Other versions5 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 5
- nahid-sparktales/agent-dispatcher★ 59same name
- markoblogo/abvx-agent-skills★ 16same name
- arslanvuzmal/VoxCircuit★ 2same name
- motao123/dev-workflow-kit★ 1same name
- SWEStash/swe-workflow-skills★ 1same name
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository28 other files it gives its agents.
- AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- .github/copilot-instructions.mdCopilot instructions
- api-designSkill
- architecture-decisionSkill
- autonomous-relaySkill
- code-review-passSkill
- data-analysisSkill
- data-modelingSkill
- data-pipelineSkill
- data-qualitySkill
- documentationSkill
- firmware-buildSkill
- frontend-buildSkill
- incident-responseSkill
- incremental-deliverySkill
- ml-lifecycleSkill
- performance-passSkill
- release-and-rollbackSkill
- requirements-interviewSkill
- security-hardeningSkill
- simplification-passSkill
- spec-writingSkill
- systematic-debuggingSkill
- tdd-loopSkill
- team-escalationSkill
- using-agent-dev-teamSkill
- work-breakdownSkill
The file
We split this skill into the rules it gives an agent
5 rules. 0 are common, 5 are distinctive, and 0 are very specific to this project.
0
Common
Most of the 4 edited versions we read keep these.
5
Distinctive
This author's own choices. Worth a look.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 20
Console: Read every error and warning, including the ones that were already there.
- Distinctive: line 23
Network: Check for a failed request, a duplicated request, a request that fires on every keystroke, and a payload much larger than the page …
- Distinctive: line 29
Keyboard only: Stop using the mouse. Tab through every control.
- Distinctive: line 30
every interactive element is reachable,
- Distinctive: line 64
Do not immediately edit. Check in this order, because the first two are free:
“Common” is measured: kept by at least 6 in 10 of the 4 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Browser Verification
A passing test says the test passes. It does not say the element is visible, that focus is where the user can see it, or that the console is full of errors. Load the page.
Sequence
- Load it. Note anything that appears wrong before you touch anything.
- Console. Read every error and warning, including the ones that were already there. A pre-existing error is still your finding; report it even if it is not yours to fix.
- Network. Check for a failed request, a duplicated request, a request that fires on every keystroke, and a payload much larger than the page uses.
- Interact. Walk the actual path a user walks. Click the primary action. Submit the form with valid input, then with invalid input, then with nothing.
- Keyboard only. Stop using the mouse. Tab through every control. Confirm:
- every interactive element is reachable,
- focus is visible at every step,
- focus order follows visual order,
- a modal traps focus, closes on
Escape, and returns focus where it came from.
- States. Force each one and look at it: empty, loading, error, partial, long content. Throttle the network to see loading; block the request to see the error. A state you never rendered is a state you have not verified.
- Responsive. 320 pixels wide, then 200% zoom. Nothing may scroll horizontally, and nothing may be clipped or overlap.
- Theme. Light and dark, if the product has both.
Evidence
Report what you observed, not what you expect:
## Browser verification: <page or component>
- **Loaded:** <url> — <viewport>
- **Console:** clean | <the errors, quoted>
- **Network:** <failed, duplicated, or oversized requests; or "clean">
- **Interaction:** <path walked, result>
- **Keyboard:** <reachable / focus visible / trap behavior>
- **States checked:** empty | loading | error | partial | long-content
- **Responsive:** 320px <result> · 200% zoom <result>
- **Findings:** <each with what a user experiences>"It renders" is not evidence. Say what you did and what happened.
When something looks wrong
Do not immediately edit. Check in this order, because the first two are free:
- Is it a stale build or a cached asset? Hard reload.
- Is the console already telling you? Read the error before reading the code.
- Is the element present but hidden, or absent? Inspect it. Those are different bugs.
- Is the data wrong, or is the rendering wrong? Check the network response.
Then, if the cause is still unknown, use systematic-debugging.
Red flags
| Thought | Reality |
| "The test passes, so the page works" | Tests do not catch invisible focus, a console full of errors, or an element behind a modal. |
| "I'll check the console if something looks broken" | Read it every time. It usually reports the bug before you see it. |
| "That console error was already there" | Report it. Pre-existing is not the same as acceptable. |
| "It looks fine on my screen" | 320 pixels, 200% zoom, long content. |
| "Keyboard access is a separate task" | It is one minute of tabbing, and it finds real bugs. |
| "I could not reproduce the loading state" | Throttle the network. Every state is forceable. |
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

